246 HOW CROPS FEED. 
surface in dry weather a portion of the ammonia which 
before was chemically retained within it. 
Solubility of the Ammonia of the Soil.—The distinc- 
tions between physically adhering and chemically combin- 
ed ammonia are difficult, if not impossible, to draw with 
accuracy. In what follows, therefore, we shall not attempt 
to consider them separately. 
When ammonia, carbonate of ammonia, or any of the 
following ammoniacal salts, viz., chloride, sulphate, ni- 
trate, and phosphate, are dissolved in water, and the solu- 
tions are filtered through or agitated with a soil, we find 
that a portion of ammonia is invariably removed from so- 
lution and absorbed by the soil. An instance of this ab- 
sorbent action has been already given in recounting 
Brustlein’s experiments, and further examples will be here- 
after adduced when we come to speak of the silicates of 
the soil. The points to which we now should direct at- 
tention are these, viz., lst, the soil cannot absorb ammo- 
nia completely from tts solutions ; and, 2d, the ammonia 
which it does absorb may be to a great degree dissolved 
out again by water. In other words, the compounds of 
ammonia that are formed in the soil, though comparatively 
insoluble, are not absolutely so. 
Henneberg and Stohmann found that a light, calcareous, 
sandy garden soil, when placed in twice its weight of pure 
water for 24 hours, yielded to the latter s35. of its weight 
of ammonia (=0.0002"|,). 
100 parts of the same soil left for 24 hours in 200 parts 
of a solution of chloride of ammonium (containing 2.182 
of sal-ammoniac =0.693 part of ammonia), absorbed 0.112 
part of ammonia. Half of the liquid was poured off 
and its place supplied with pure water, and the whole 
left for 24 hours, when half of this liquid waa taken, and 
the process of dilution was thus repeated to the fifth time. 
In the portions of water each time removed, ammonia was 
estimated, and the result was that the water added dis- 
